TWIST2 and the PPAR Signaling Pathway Are Important in the Progression of Nonalcoholic Steatohepatitis

Zhang et al. Lipids in Health and Disease (2021) 20:39 https://doi.org/10.1186/s12944-021-01458-0

RESEARCH Open Access TWIST2 and the PPAR signaling pathway are important in the progression of nonalcoholic steatohepatitis Yanmei Zhang1,2†, Xiaoxiao Ge1†, Yongqing Li1, Bingyang Zhang1, Peijun Wang1, Mingju Hao1,3, Peng Gao4, Yueyi Zhao3, Tao Sun1,3, Sumei Lu1,3* and Wanshan Ma1,3*

Abstract Background: To investigate the roles of the transcription factors twist family bHLH transcription factor 1 (TWIST1), twist family bHLH transcription factor 2 (TWIST2), and peroxisome proliferator activated receptor gamma (PPARγ)in the progression of nonalcoholic steatohepatitis. Methods: The protein levels of TWIST1, TWIST2 and PPARγ were determined in the serum of nonalcoholic fatty liver disease (NAFLD) patients and healthy controls by enzyme-linked immunosorbent assay (ELISA). An in vivo model for fatty liver was established by feeding C57BL/6 J mice a high-fat diet (HFD). An in vitro model of steatosis was established by treating LO-2 cells with oleic acid (OA). RNA sequencing was performed on untreated and OA- treated LO-2 cells followed by TWIST1, TWIST2 and PPARγ gene mRNA levels analysis, Gene Ontology (GO) enrichment and pathway analysis. Results: The TWIST2 serum protein levels decreased significantly in all fatty liver groups (P < 0.05), while TWIST1 varied. TWIST2 tended to be lower in mice fed an HFD and was significantly lower at 3 months. Similarly, in the in vitro model, the TWIST2 protein level was downregulated significantly at 48 and 72 h after OA treatment. RNA sequencing of LO-2 cells showed an approximately 2.3-fold decrease in TWIST2, with no obvious change in TWIST1 and PPARγ. The PPAR signaling pathway was enriched, with 4 genes upregulated in OA-treated cells (P = 0.0018). The interleukin (IL)-17 and tumor necrosis factor (TNF) signaling pathways were enriched in OA-treated cells. Conclusions: The results provide evidence that the TWIST2 and PPAR signaling pathways are important in NAFLD and shed light on a potential mechanism of steatosis. Keywords: Nonalcoholic fatty liver disease, Hepatocyte steatosis, TWIST family bHLH transcription factor 1 (TWIST1), TWIST family bHLH transcription factor 2 (TWIST2), Peroxisome proliferator activated receptor gamma (PPARγ), The PPAR signaling pathway, High-fat diet

* Correspondence: [email protected]; [email protected] †Yanmei Zhang and Xiaoxiao Ge contributed equally to this work. 1Department of Laboratory Medicine, Shandong Provincial Qianfoshan Hospital, Shandong University, Jinan, Shandong 250014, P. R. China Full list of author information is available at the end of the article

© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 2 of 13

Introduction to ameliorate hepatocellular steatosis and inhibit inflamma- Nonalcoholic fatty liver disease (NAFLD) is a common tion [16]. chronic liver disorder characterized by lipid accumulation The present study will continue previous work to in hepatocytes. Currently, NAFLD affects approximately explore the role and possible mechanism of TWIST1, 20–30% of the Western population and 5–18% of the TWIST2, and PPARγ in the development of hepatocyte Asian population and is continuing to increase [1]. More- steatosis. This work will mainly be based on clinical over, being a metabolic disease, NAFLD can aggravate samples and in vivo and in vitro models of hepatocyte other metabolic syndromes, including obesity, type 2 steatosis. The results will provide evidence for the role diabetes mellitus, dyslipidemia, hypertension and insulin of TWIST and PPARγ in NAFLD, clarify their function resistance (IR) [2, 3]. However, the pathogenesis and mo- in the adipogenesis process and provide new insights lecular mechanism of NAFLD are still unclear, and further into steatosis-related diseases. work is undoubtedly necessary. NAFLD includes simple steatosis as well as nonalco- Materials and methods holic steatohepatitis (NASH), a more serious condition Patient inclusion and human serum sample collection with associated inflammation. It is well known that A total of 406 individuals who had a health check-up NASH can progress to more serious conditions, such as at Shandong Provincial Qianfoshan Hospital from fibrosis, cirrhosis and hepatocellular carcinoma [4]. The December 2017 to December 2018 were recruited into the widely accepted pathogenesis of NAFLD is the “multiple present study. Following the inclusion criteria, these hit theory”, which suggests that in the development of NAFLD patients were generally 18–60 years old, had nor- NAFLD, many hits might work together in parallel mal thyroid function, no history of drinking, no malig- [5–7]. Obesity and IR are reported to play a fundamental nancy, no viral hepatitis, no drug-induced liver disease, no role in the progression of NAFLD. High insulin levels pro- autoimmune liver disease, and all other specific diseases mote lipid synthesis by increasing glycolysis in the liver that cause fatty liver. [8]. In addition, high insulin also reduces the expression of NAFLD patients were divided into mild, moderate, ApoB-100 and microsomal triglyceride transfer protein and severe NAFLD groups based on abdominal ultra- (MTTP), leading to a reduction in very low density lipo- sound testing. The standard of ultrasound fatty liver protein (VLDL) output in the liver, eventually leading to classification is mainly based on the shape of the liver, lipid deposition in the liver tissue [9]. IR can also cause the echo of the liver parenchyma, the degree of attenu- hepatic steatosis, and it has been proposed that hepatic ation behind the liver, and the visualization of the intra- steatosis further aggravates hepatic insulin resistance, hepatic duct. Generally, patients with mild fatty liver creating a feedback loop between the two conditions [10]. have a normal liver shape and a sharp angle. In patients Kim et al. suggested that hepatic steatosis could lead to with moderate fatty liver, the liver is slightly larger, and IR, but some scholars believe that hepatic steatosis is not the angle becomes blunt. In patients with severe fatty enough to cause liver IR [11, 12]. Recently, much more is liver, the liver is obviously enlarged, and the angle is said about molecular mechanisms because the mecha- round and blunt. The liver parenchyma echoes, the nisms that lead to the release of many mediators are patients with mild fatty liver are fine and thicker, and known, which then lead to the formation of proliferation the echoes of patients with severe fatty liver are obvi- and scar tissue after a long inflammatory process, induc- ously thicker. The attenuation at the rear of the liver tion of apoptosis and/or necrosis [13]. The relationship changes from normal to obvious, and the intrahepatic between hepatic steatosis and IR is still controversial and ducts change from normal to unclear. Different grades needs further study. can distinguish mild, moderate and severe fatty liver. The authors’ research team has previously reported The condition was diagnosed by the ultrasound physician. that two transcription factors, TWIST1 and PPARγ, Healthy adults who were also having a health check-up at have a positive regulatory role in the insulin sensitivity the same time had samples randomly collected. This study of 3 T3-L1 adipocytes [14]. In IR models of 3 T3-L1 was performed in accordance with ethical standards adipocytes and C57/BL6J mice, silencing TWIST1 and was approved by the Ethics Committee of Shandong expression can relieve IR to a certain degree, indicating University (No: [2017] S048). the potential clinical value of TWIST1 and PPARγ in steatosis-related disease [15]. As hepatocytes are target Materials and reagents cells of insulin, the role of TWIST1 and PPARγ in hepa- C57/BL6 mice were purchased from the Experimental tocytes is undoubtedly worth further exploration. At the Animal Center of Shandong University. The high-fat diet same time, TWIST2, the other member of the TWIST (No: D12492) and basal diet (No: D12450B) were both family, attracted attention for its role in metabolism and products of Research Diets (New Brunswick, USA). The metabolic diseases. Overexpression of TWIST2 was reported human hepatic cell line LO-2 (also named HL-7702) was Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 3 of 13

purchased from Procell Life Science & Technology Co., Intraperitoneal glucose tolerance test (IPGTT) and Ltd. (Wuhan; China). The human TWIST1, TWIST2 intraperitoneal insulin tolerance test (IPITT) and PPARγ ELISA kits were Jianglai Biology (Shanghai, Before the mice were sacrificed at the 16th week, IPGTT China) products. RPMI-1640 medium, fetal bovine and IPITT were conducted in a portion of the mice to serum (FBS) and 0.25% trypsin-0.02% EDTA were all determine the ability of mice to respond to glucose and purchased from GIBCO (Invitrogen, California, USA). insulin. After fasting overnight for IPGTT or six hours The rabbit anti-mouse TWIST1 mAb was purchased for IPITT, 12 mice were injected with 50% glucose (2.0 from Sigma (St. Louis, MO, USA). Rabbit anti-rabbit g/kg, i.p.), and 12 mice were injected with insulin (0.65 TWIST2 pAb and rabbit anti-mouse PPARγ mAb were U/kg, i.p). Blood glucose levels in the mice were ana- purchased from Abcam (Cambridge, MA, USA). The lyzed by means of a glucose test strip (Bayer, Germany) rabbit anti-mouse β-actin mAb and secondary antibodies, after blood samples were taken via the caudal vein at 0, including horseradish peroxidase-conjugated anti-mouse 30, 60, 90, 120, 150 and 180 min after glucose or insulin IgG for TWIST1, PPARγ,andβ-actin and horseradish injection. The area under the curve (AUC) was com- peroxidase-conjugated anti-rabbit IgG for anti-TWIST2, pared among different groups to analyze the IPGTT and were purchased from ZSGB-BIO (Beijing, China). Oleic IPITT. acid (OA) and all other general reagents used in this study were purchased from Sigma (St. Louis, MO, USA). Establishment of the LO-2 hepatocyte steatosis model LO-2 hepatocytes were cultured in RPMI 1640 supplemented with 10% fetal bovine serum, 100 U/mL Serum determination of TWIST1, TWIST2 and PPARγ in penicillin and 100 μg/mL streptomycin. The cells were patients cultured in a humidified chamber at 37 °C and 5% The levels of TWIST1, TWIST2 and PPARγ in the CO . For steatosis induction, RPMI-1640 supplemented serum of patients were analyzed by ELISA per the 2 with 10% FBS and OA (50 μg/mL) was added to the cells manufacturer’s instructions. Generally, standard or when they reached approximately 60% confluence. The serum (each 50 μL) was added to the plate wells; culture medium and OA were replaced every 24 h for 24, HRP-conjugated antibody (100 μL) was added to the 48, 72, and 96 h. plate to mix with standard or serum at 37 °C for 1 h. After being washed 5–6 times, the plates were treated Hematoxylin-eosin (H&E), oil red O staining of frozen liver with 50 μL each of substrates A and B and incubated sections and LO-2 hepatocytes at 37 °C for 15 min, and then 50 μL of terminating After the mice were sacrificed and the livers were re- buffer was added to stop the reaction. Finally, the OD moved and snap frozen, H&E staining was performed on value was checked under an enzyme-linked immuno- frozen sections. The general steps are as follows: Metha- assay analyzer at 450 nm. nol fixative for 60 s; Hematoxylin 2 min; Washed with phosphate buffered saline (PBS) for 30 s; Hydrochloric Induction of NAFLD in C57/BL6 mice acid alcohol washed for 3 s; Washed with PBS for 30 s; Sixty-four male C57/BL6 mice (6 weeks of age) were Lithium carbonate 10 s; Washed with PBS for 30 s; 90% included in the study and housed 6 per cage. The mice alcohol for 10 s; Eosin for 5 s; 90% alcohol for 5 s; 95% were housed in a specific environment: 25 °C, 55% alcohol for 20 s, twice; 100% alcohol for 20 s; 100% relative humidity, cycle of 12 h light and 12 h dark. They alcohol for 20 s; Xylene for 30 s, twice. The resulting were free to eat and drink tap water ad libitum. After a slides were visualized with an inverted phase contrast week, they were randomly assigned to two groups, the microscope. control group (basal diet, 4% fat, with 20 g% lard) and For Oil red O staining, LO-2 cells were grown on the high-fat diet (HFD) group (HFD, 60% fat, with 225 cover slips, and after removal of cell culture medium, g% lard). They were fed a control diet or HFD for 4, 8, cells were washed three times with PBS and fixed with 12, or 16 weeks, and their body weights were recorded 4% tissue cell fixative solution (Solarbio Science) at room every week. After the mice were euthanized, liver tissues temperature for 10 min. Frozen slices for hepatic tissue were collected and snap frozen. Hematoxylin and eosin were similarly fixed. After washing three times with PBS, (H&E) staining and Oil red O staining were conducted the cells or frozen slices were stained with freshly diluted on liver sections. The additional procedures were carried Oil red O solution (0.1% Oil red O dissolved in 60% iso- out according to the ‘Principles of Laboratory Animal propyl alcohol and 40% distilled water) for 15–30 min. Care’ established by the National Institutes of Health. Cells or frozen slices were then washed with 60% isopro- This study was performed in accordance with ethical pyl alcohol for 1 min and washed with PBS three times. standards and was approved by the Ethics Committee of Images were obtained using an inverted phase contrast Shandong University (No: [2017] S048). microscope. Oil red O staining was quantified by an Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 4 of 13

absorbance assay. Briefly, the Oil red O stain was solubi- Statistical analysis lized with isopropyl alcohol and the optical density at The data were analyzed using SPSS 24.0 software 510 nm was measured by spectrophotometry (Multiskan (mean ± SD). The correlations between TWIST1, TWIS Go, Thermo Scientific). All experiments were conducted T2, PPARγ, body mass index (BMI), fasting glucose, fast- in triplicate. ing insulin, or homeostasis model assessment (HOMA) and fatty liver severity were analyzed by ANOVA. For Protein extraction and western blotting analysis western blotting, we used Image J to determine the rela- Total protein was extracted using radioimmunopreci- tive gray values of each band by comparison with β- pitation assay (RIPA) lysis buffer containing protease actin. A t test was used to compare the differences be- inhibitor cocktail (Thermo Scientific, USA) and phos- tween two groups, and ANOVA was used to compare phatase inhibitor cocktail (Thermo Scientific, USA), the differences between more than two groups. P < 0.05 and the protein concentration was analyzed by Pierce™ was considered an indicator of significant difference. BCA protein assay as previously shown [14, 15]. The protein (40 μg for each sample) was resolved on SDS- Results PAGE gels and transferred to a Hybond-P PVDF mem- Analysis of TWIST1, TWIST2, and PPARγ protein levels in brane. The membranes were blocked with 5% milk or human serum samples 5% bovine serum albumin (BSA) for 1 h. The mem- The serum samples were collected at the clinical labora- branes were incubated with primary antibodies (anti- tory. A total of 406 human serum samples were col- TWIST1, 1:1000; anti-TWIST2, 1:1000; anti-PPARγ,1: lected in this study. The samples were divided into four 1000 and anti-β-actin, 1:2500) at 4 °C overnight. After groups based on abdominal ultrasound testing, with a washing three times (20 min each) with TBST, the sec- healthy control group (n = 99), mild NAFLD group (n = ondary antibodies were incubated with horseradish 102), moderate NAFLD group (n = 107), and severe peroxidase-conjugated anti-rabbit or anti-mouse (1: NAFLD group (n = 98). The Materials and Methods sec- 5000) antibodies for 1 h at room temperature. Blots tion describes the criteria for determining whether a were washed three times (20 min each), and image de- fatty liver is mild, moderate, or severe. Sex, age, height, velopment was performed with electrochemilumines- and body weight were recorded for each individual, and cence (ECL) reagent. All experiments were performed the basic clinical characteristics of these individuals are in triplicate. shown in Table 1. They were further divided into four

RNA sequencing analysis Table 1 Basic information on the human serum samples (n = RNA sequencing analysis was performed using the 406). A total of 406 human serum samples were collected. Each Illumina HiSeq 4000 platform (KangChen Bio-tech, individual’s sex, age, height, and body weight were recorded Shanghai, China). Briefly, cultured LO-2 cells were and are shown in Table 1 treated with DMSO (the control group) and OA Parameter Number (50 μg/mL, 48 h). After treatment, cells were lysed with Sex TRIzol reagent. RNA was extracted as normal proto- male 344 col. RNA (1–2 μg) was used to establish the RNA library based on specific kits. It includes RNA fragmenta- female 62 tion, primed 1st strand cDNA synthesis, dUTP-based 2nd Age (years) strand cDNA synthesis, adaptor ligation and PCR amplifi- 18–44 301 cation. The library was qualified using an Agilent 2100, 45–59 105 quantified by qPCR, and then subjected to sequencing BMI (kg/m2) based on Illumina X-ten/NovaSeq. Three biological repli- < 18.5 4 cates were prepared in each group. Genes were considered differentially expressed if the expression difference be- 18.5–23.9 84 tween two groups showed a twofold change or greater 24–27.9 137 (log2FC > 1.0, fold change). Gene Ontology (GO) enrich- ≥ 28 181 ment analysis was performed on the http://geneontology. Fatty liver org website, and the enrichment scores for molecular healthy control 99 function (MF), biological process (BP) and cellular compo- mild fatty liver 102 nent (CC) were analyzed. Pathway analysis for differentially expressed genes was further conducted based on the moderate fatty liver 107 Kyoto Encyclopedia of Genes and Genomes (KEGG) data- severe fatty liver 98 base, whose website is https://www.genome.jp/kegg/. BMI body mass index Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 5 of 13

groups based on BMI (kg/m2), with thin group (BMI< The increased cell density may be the possible cause of 18.5 kg/m2, n = 4), normal group (BMI = 18.5–23.9 kg/ TWSIT1 and PPARγ changes in LO-2 control cells m2, n = 84), overweight group (BMI = 24–27.9 kg/m2, under no external stimulation. Statistical analysis based n = 137), and obesity group (BMI ≥ 28 kg/m2, n = 181). on semi-quantitative analysis using Image J software As shown in Table 2, fasting blood glucose and insulin showed that Both TWIST1 (Fig. 2d, the arrow showed levels were determined by an automatic analyzer. Both the true band) and TWIST2 (Fig. 2e) in OA treatment blood glucose and blood insulin increased significantly tended to be decreased, and a significant difference with the severity of fatty liver disease (P < 0.05). Spear- could be found in TWIST2 at both 48 and 72 h. No sig- man rank correlation analysis showed that a positive nificant change in PPARγ (Fig. 2f) expression between correlation was found between the severity of fatty liver the control and OA (50 μg/mL) treatment groups disease and BMI (r = 0.761, P < 0.001) and HOMA value (P > 0.05). (r = 0.607, P < 0.001). Then, TWIST1, TWIST2 and PPARγ levels were measured in the serum of patients RNA sequencing analysis and RT-PCR both showed that using ELISA kits. For TWIST1, the levels in the mild TWIST2 was decreased significantly by OA treatment and severe fatty liver groups decreased significantly △ RNA sequencing of LO-2 cells treated with OA for ( P < 0.05, compared with healthy control) and increased △ 48 h showed that 1884 genes were upregulated and significantly in the moderate fatty liver group ( P < 0.05, 1975 genes were downregulated in the OA treatment compared with healthy control). The levels of TWIST2 group compared with the control group (Fig. 3a). in all of the fatty liver groups decreased significantly △ Cluster analysis of the top 20 upregulated and down- ( P < 0.05, compared with healthy control). For PPARγ, regulated genes is shown in Fig. 3b. TWIST1 and compared with the control group, there were no signifi- PPARγ mRNA levels showed no significant difference cant changes in the mild fatty liver, moderate fatty liver, (P > 0.05) between the groups in the RNA sequencing and severe fatty liver groups (P > 0.05). results, but TWIST2 was downregulated significantly with OA treatment (P <0.05) (Fig. 3c). Further GO Characterization of the in vivo model of NAFLD enrichment analysis of the RNA-seq data revealed 542 After 16 weeks on a high-fat diet, the livers in the HFD differentially expressed (DE) upregulated genes and group were enlarged and had a white hue (Fig. 1a). H&E their associated biological processes (BPs) (Fig. 3d) staining (Fig. 1b) and Oil red O staining (Fig. 1c) showed and 514 DE downregulated genes and associated BPs increased lipid and triglyceride contents in the HFD (Fig. 3e). The details of the top 20 upregulated and group indicative of steatosis. The mice in the HFD group downregulated genes under oleic acid treatment by had a much higher body weight that increased signifi- RNAsequencingareshowninTable3. cantly from 6 weeks of age (*P < 0.05, Fig. 1d). Both the IPGTT (Fig. 1e) and IPITT (Fig. 1f) of the HFD group showed a decreased insulin response. TWIST1, TWIST2 Genes in the PPAR signaling pathway were enriched and PPARγ protein levels in liver tissues all changed significantly based on RNA sequencing with feeding time (Fig. 1g). Both TWIST1 (the arrow Compared with the control, the DE genes in OA- showed the true band) and TWIST2 expressions in HFD treated LO-2 cells (50 μg/mL, 48 h) were involved in 62 tended to be decreased in the detection, and statistical pathways, with 42 upregulated (top 10 shown in Fig. 4a) significance showed only in TWIST2 at 3 months after and 20 downregulated (top 10 shown in Fig. 4b). In the the start of the HFD (Fig. 1h, i). There were no signifi- top 10 upregulated pathways of DE genes, the IL-17 cant change in PPARγ expression (Fig. 1j). signaling pathway [hsa04657], TNF signaling pathway [hsa04668], and PPAR signaling pathway [hsa03320] Characterization of the in vitro model of hepatocyte were listed. Four genes in the PPAR signaling pathway steatosis were upregulated in OA (50 μg/mL, 48 h)-treated cells LO-2 cells were treated with OA (50 μg/mL) for 24, 48, (P = 0.0018). These four genes are acyl-CoA synthetase 72, and 96 h to induce steatosis. Obvious red-stained long chain family member 4 (ACSL4), angiopoietin like lipid droplets could be seen with Oil red O staining 4 (ANGPTL4), carnitine palmitoyltransferase 1A (CPT1A), compared with the control (Fig. 2a, b), showing that the and perilipin 4 (PLIN4). RT-PCR verified their up regula- LO-2 cell steatosis model was established. tion (Fig. 4c). A schematic model for the PPAR function is As shown in Fig. 2c, both TWIST1 (the arrow showed proposed, as shown in Fig. 4d. During adipogenesis, PPAR the true band) and PPARγ protein expression gradually can be induced by unsaturated fatty acids or 9-cis-retinoic increased with time (24, 48, 72, and 96 h) under OA acid and can upregulate ACSL4, ANGPTL4, CPT1A, treatment in LO-2 cells. However, similar increases and PLIN, which can regulate lipid metabolism or could be found in control cells without OA exposure. adipocyte differentiation. Zhang ta.Lpd nHat n Disease and Health in Lipids al. et (2021)20:39

Table 2 Data analysis for the human serum samples. The levels of BMI, glucose, insulin, HOMA value, Twist1, Twist2, and PPARγ were recorded. Statistical analysis was performed using ANOVA. P < 0.05 was considered to indicate a significant difference − Parameter BMI (X ± SD) glucose insulin HOMA ( X ± SD) Twist1 concentration Twist2 PPARγ − − − (mmol/L, X ± SD) (mIU/L, X ± SD) (pg/mL, X ± SD) concentration concentration Group − − (pg/mL, X±SD) (mmol/L, X ± SD) □ ☆ ※ □ ☆ ※ □ ☆ ※ □ ☆ ※ □ ☆ healthy control 22.42 ± 2.54 , , 4.98 ± 0.50 , , 6.93 ± 3.12 , , 1.54 ± 0.74 , , 287.54 ± 264.02 , 2.63 ± 2.26 258.88 ± 241.26 △ ☆ ※ △ ☆ ※ △ ☆ ※ △ ☆ ※ △ ☆ ※ △ mild fatty liver 26.32 ± 2.45 , , 5.35 ± 0.50 , , 9.19 ± 4.18 , , 2.19 ± 1.02 , , 176.62 ± 166.08 , , 2.00 ± 1.34 227.84 ± 208.65 △ □ ※ △ □ ※ △ □ ※ △ □ ※ △ □ ※ △ moderate fatty liver 28.25 ± 2.75 , , 5.92 ± 1.59 , , 11.82 ± 6.77 , , 3.15 ± 2.07 , , 380.39 ± 331.48 , , 2.04 ± 1.37 216.30 ± 177.04 △ □ ☆ △ □ ☆ △ □ ☆ △ □ ☆ □ ☆ △ severe fatty liver 31.30 ± 3.31 , , 6.29 ± 1.93 , , 15.29 ± 7.66 , , 4.20 ± 2.23 , , 279.26 ± 238.77 , 2.06 ± 1.16 225.72 ± 182.68 △ □ ✩ ※ BMI: body mass index, kg/m2, HOMA homeostasis model assessment; P < 0.05, compared with healthy controls; P < 0.05, compared with mild fatty liver; P < 0.05, compared with moderate fatty liver; P < 0.05, compared with severe fatty liver ae6o 13 of 6 Page Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 7 of 13

Fig. 1 (See legend on next page.) Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 8 of 13

(See figure on previous page.) Fig. 1 Characterization of an in vivo model of fatty liver disease and changes in TWIST1, TWIST2 and PPARγ protein expression in the liver (n = 8). a Comparison of the sizes of control and high-fat diet (HFD) group mice and their livers. b H&E staining of livers from control and HFD mice.c Oil red O staining of livers from control and HFD group mice. d Changes in body weight over time. e Intraperitoneal glucose tolerance test (IPGTT). f Intraperitoneal insulin tolerance test (IPITT). g Relative protein levels of TWIST1, TWIST2 and PPARγ in the livers of control and HFD group mice. The arrow shows that the upper band is the true Twist1. H/I/J) Semiquantitative assays of TWIST1 (h), TWIST2 (i), and PPARγ (j) in the livers of control and HFD group mice

Discussion article showed that TWIST2 was involved in the develop- The present study focuses on the role of TWIST1, ment of NAFLD [16]. TWIST2 is a basic helix-loop-helix TWIST2 and PPARγ in nonalcoholic steatohepatitis pro- (bHLH) transcription factor that is highly related to TWIS gression. To do this, the serum protein levels of TWIS T1 and appears to have some redundant and nonredun- T1, TWIST2, and PPARγ in the serum of NAFLD pa- dant functions [17]. TWIST transcription factors bind as tients were examined with varying stages of disease, and homodimers or heterodimers to E-box consensus sites significant changes in the TWIST proteins were noted. and are transcriptional activators or repressors depending In vitro and in vivo models of the disease were also stud- on the cellular context [17, 18]. TWIST proteins have ied, and TWIST2 seemed to be affected the most in been extensively studied for their oncogenic properties these models. The results using each of these models and their role in the epithelial-to-mesenchymal transition provide additional evidence that the TWIST2 transcrip- (EMT), but their involvement in other processes is less tion factor may play a role in NAFLD. well known [19]. twist2-null mice have many abnormal- TWIST2 seems to be particularly important in NAFL ities in fat and glucose metabolism and are presumed to D because the results and the publication of a recent die soon after birth because of the upregulation of

Fig. 2 The LO-2 steatosis model shows increased lipid uptake and decreased TWIST2 expression. a Oil red O staining in the control and oleic acid (50 μg/mL) groups. b OD values for oil red O staining in the control and oleic acid groups. c TWIST1, TWIST2 and PPARγ expression in the control and oleic acid treatment groups in cultured LO-2 cells. β-Actin was used as a loading control. The arrow shows that the upper band is the true Twist1. d/e/f: Semiquantitative assays of TWIST1 (d), TWIST2 (e), and PPARγ (f) Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 9 of 13

Fig. 3 (See legend on next page.) Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 10 of 13

(See figure on previous page.) Fig. 3 RNA sequencing analysis of transcriptional changes and GO biological process analysis in cultured LO-2 cells treated with OA (50 μg/mL, 48 h). a After sequencing analysis, the mRNA transcriptional changes between the control and test groups were analyzed. Shown here is a scatter map of mRNA transcriptional changes. b The top 20 upregulated and downregulated genes in the OA treatment group were selected and are shown in a heat map. c RT-PCR was conducted to determine TWIST1, TWIST2 and PPARγ mRNA changes in response to OA treatment as described in the methods section. The results are shown in the bar graphs. d GO biological process analysis was conducted, and the classifications of the upregulated genes are shown. e The classifications of the downregulated genes for GO biological processes are shown cytokines in an NF-κB-dependent manner [20]. A de- changes in TWIST2 protein content can be indicative of a crease in TWIST2 protein in the in vitro and in vivo developing disease process. models of NAFLD was found. Interestingly, a decrease in TWIST1 serum protein levels varied greatly with the TWIST2 protein levels in the serum of NAFLD patients at severity of disease, with significantly lower levels in the all stages of the disease was also noticed. This suggests mild fatty liver group and significantly higher levels in that in NAFLD, a reduction in the level of TWIST2 pro- the moderate fatty liver group. The level of TWIST1 tein may be an early event that increases inflammation. It protein in the severe fatty liver group was the same as also suggests that TWIST2 protein levels in serum may be that in the control group. This pattern of TWIST1 a biomarker of NAFLD. It has recently been shown that serum protein may indicate complex regulation of the TWIST1 mRNA and protein can be found in exosomes TWIST1 gene during the development of NAFLD. secreted from hepatic stellate cells (HSCs) and that the TWIST1 was considered an anti-inflammatory marker, levels of TWIST1 mRNA and protein decrease in models as low mRNA and protein expression of TWIST1 was of hepatic injury and fibrosis [21]. It is likely that TWIST2 associated with increased expression of proinflammatory is also secreted from the liver compartment and that cytokines and decreased insulin sensitivity in humans

Table 3 Details on the top 20 upregulated and downregulated genes under oleic acid treatment as determined by RNA sequencing. The top 20 upregulated genes are listed, including, among others, ASNS-209, APLP2–202, and PSAP-201. The top 20 downregulated genes included, among others, APLP2–201, DDX27–206, and CARD10–203. All the genes were significantly differentially expressed (P < 0.05)

Gene name Transcript log2(fold Fold P value Gene Transcript log2(fold Fold P value name change) change name name change) change Upregulated genes Downregulated genes ASNS ASNS-209 4.9946 31.8794 0.0000 APLP2 APLP2–201 −3.2019 0.1087 0.0001 APLP2 APLP2–202 4.0958 17.0984 0.0070 DDX27 DDX27–206 − 3.1990 0.1089 0.0193 PSAP PSAP-201 4.0106 16.1183 0.0283 CARD10 CARD10–203 −2.9708 0.1276 0.0235 EIF4G1 EIF4G1–221 3.7234 13.2089 0.0349 JADE3 JADE3–203 − 2.7405 0.1496 0.0000 ANGPTL4 ANGPTL4–202 3.5879 12.0241 0.0008 APOL2 APOL2–201 −2.7045 0.1534 0.0246 NDUFS8 NDUFS8–210 3.3598 10.2658 0.0235 POLR2A POLR2A-201 −2.5492 0.1709 0.0266 POLR2B POLR2B-213 3.3478 10.1812 0.0213 KIF14 KIF14–202 −2.5153 0.1749 0.0272 SERPINE2 SERPINE2–203 3.3009 9.8552 0.0139 CD59 CD59–203 −2.4796 0.1793 0.0002 HAX1 HAX1–202 3.1570 8.9197 0.0238 KATNA1 KATNA1–205 −2.3728 0.1931 0.0247 CCDC47 CCDC47–207 3.1261 8.7309 0.0229 EIF4G1 EIF4G1–216 −2.3661 0.1940 0.0057 TIPARP TIPARP-206 3.0047 8.0262 0.0214 POLR2B POLR2B-205 −2.3462 0.1967 0.0220 ANGPTL4 ANGPTL4–201 2.9858 7.9219 0.0001 ATP5G2 ATP5G2–204 −2.3328 0.1985 0.0145 UBA2 UBA2–208 2.8114 7.0199 0.0272 NOP2 NOP2–203 −2.3103 0.2016 0.0275 EIF4G1 EIF4G1–207 2.6434 6.2481 0.0280 GPANK1 GPANK1–201 −2.2093 0.2162 0.0022 PPP2R2A PPP2R2A-201 2.5557 5.8797 0.0221 SEC22C SEC22C-205 −2.1893 0.2193 0.0258 PDK4 PDK4–201 2.5308 5.7790 0.0000 ALPI ALPI-201 −2.1868 0.2196 0.0045 ZFAND2B ZFAND2B-206 2.4906 5.6202 0.0000 CTSA CTSA-201 −2.1791 0.2208 0.0134 DDX19B DDX19B-208 2.4219 5.3586 0.0370 ALDH3B1 ALDH3B1–208 −2.1746 0.2215 0.0139 ZYX ZYX-203 2.4101 5.3151 0.0480 MEOX1 MEOX1–202 −2.1356 0.2276 0.0000 WASHC5 WASHC5–202 2.2769 4.8462 0.0226 SNX3 SNX3–204 −2.1237 0.2295 0.0156 Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 11 of 13

Fig. 4 PPAR signaling pathway genes were enriched significantly in LO-2 cells by OA treatment (50 μg/mL, 48 h). a KEGG pathway analysis based on the KEGG website. The top 10 upregulated pathways of differentially expressed (DE) genes are shown here. b The top 10 downregulated pathways of DE genes based on the KEGG website shown here. c The mRNA levels of four genes in the PPAR signaling pathway, namely, ACSL4, ANGPTL4, CPT1A, and PLIN4, showed increased transcription based on RT-PCR. d A schematic model of the PPARγ pathway in lipid metabolism and adipocyte differentiation was drawn, indicating the mechanism of the PPAR signaling pathway in NAFLD

[22]. Lower levels of TWIST1 in the early stages of PPAR signaling pathway was involved in the hepatic NAFLD may lead to increased inflammation and disease steatosis process. It is believed that PPAR may regulate progression. downstream lipid metabolism or adipocyte differenti- Increased protein expression of PPAR is a general ation through ACSL4, ANGPTL4, CPT1A, and PLIN4. property of steatotic livers. The contribution of PPAR is Further investigation into the mechanism of how PPAR to the maintenance of a steatotic phenotype in liver cells contributes to NAFLD is warranted. [23]. PPARγ is capable of activating the expression of Chronic inflammation has also been implicated in the genes involved in TG accumulation in hepatocytes and development and progression of NAFLD [25, 26]. TWIS promoting the generation of fatty liver [24]. While no T1 and TWIST2 have been shown to reduce inflamma- change was seen in PPARγ protein levels in the patient tion by inhibiting the NF-kB pathway [18, 20]. As men- serum or in the in vivo and in vitro models, the RNA se- tioned, twist-2-deficient mice produce high levels of quence analysis of OA-treated cells suggested that the cytokines [20]. An interesting finding of the GO analysis Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 12 of 13

of the RNA sequencing data revealed an enhanced IL-17 Acknowledgments signaling pathway in OA-treated cells. IL-17 signaling RNA sequencing analysis was performed by KangChen Bio-tech, Shanghai, China. has become of greater significance in NAFLD in recent years [27]. An enhancement in the TNF signaling path- Authors’ contributions way was also found, another important inflammatory SML and WSM contributed to funding acquisition, study concept, study – design, data interpretation and revision of the manuscript. YMZ and XXG pathway implicated in NAFLD [28 30]. It has been re- performed most of the experiments and contributed to data collection, data ported that a significantly higher concentration of TNF- analysis and original draft writing; YQL, MJH, and YYZ contributed to data alpha is associated with both the type of inflammatory analysis; BYZ, PG, and TS performed some part of the research. All authors cells and with increased vascularization in various tissue read and approved the final manuscript. types [31]. Funding This study was funded by the Cultivation Fund of National Natural Science Foundation of China in Shandong Provincial Qianfoshan Hospital (No. Study strength and limitations QYPY2020NSFC1004), the Key Technology Research and Development The present findings prompt TWIST2 as a potential Program of Shandong (No. 2017G006024), National Natural Science Foundation of China (No. 81400843), and Natural Science Foundation of new biomarker for the early detection of NAFLD. Mean- Shandong Province (No. ZR2014HP033). while, the involvement of the PPAR signaling pathway hints further perspective in hepatocytes steatosis. How- Availability of data and materials ever, there are some limitations in this study. Although The data will be available on request. the function of PPAR is involved in steatosis, the mech- Declarations anism of how PPAR contributes to NAFLD is unclear. It is necessary to determine the PPAR gene function and Ethics approval and consent to participate This study was performed in accordance with ethical standards and was to study the phenotype if we block the PPAR signaling approved by the Ethics Committee of Shandong University (No: [2017] S048). pathway. Also, the inconsistence of TWIST1 gene Informed consent was obtained from all individual participants included in expression at protein level between in vitro, in vivo and the study. in patient samples is needed to be fully examined. The Consent for publication mechanism of downregulation of the TWIST2 gene ex- Not applicable. pression in 3rd month is also needed to be understood. Competing interests Together this paper demonstrates the importance of the The authors declare that they have no conflicts of interest. TWIST2 gene in the clinical prognosis and roles of PPAR and TWIST1 in NAFLD development needed Author details 1Department of Laboratory Medicine, Shandong Provincial Qianfoshan fully investigated. Hospital, Shandong University, Jinan, Shandong 250014, P. R. China. 2Department of Clinical Laboratory, the Affiliated Hospital of Shandong University of Traditional Chinese Medicine, Jinan, Shandong 250014, P.R. Conclusions China. 3Department of Laboratory Medicine, The First Affiliated Hospital of In summary, the present study shows that the TWIST2 Shandong First Medical University, Jinan, Shandong 250014, P. R. China. 4 and PPAR signaling pathways are important in NAFLD Medical Research Center, The First Affiliated Hospital of Shandong First Medical University, Jinan, Shandong 250014, P. R. China. and sheds light on a potential mechanism of steatosis. The present findings establish TWIST2 as a new Received: 17 November 2020 Accepted: 29 March 2021 biomarker for the early detection of NAFLD. A simple blood test for TWIST2, a potential biomarker, may allow References one to detect the early stages of NAFLD. Future studies 1. Povero D, Feldstein AE. Novel molecular mechanisms in the development on the PPAR signaling pathway will be urgent to clarify of non-alcoholic Steatohepatitis. Diabetes Metab J. 2016;40(1):1–11. https:// doi.org/10.4093/dmj.2016.40.1.1. the roles of these genes in NAFLD. 2. Bellentani S. The epidemiology of non-alcoholic fatty liver disease. Liver Int. 2017;37(Suppl 1):81–4. https://doi.org/10.1111/liv.13299. Abbreviations 3. Watanabe S, Hashimoto E, Ikejima K, Uto H, Ono M, Sumida Y, et al. TWIST1: Twist family bHLH transcription factor 1; TWIST2: Twist family bHLH Evidence-based clinical practice guidelines for nonalcoholic fatty liver transcription factor 2; PPARγ: Peroxisome proliferator activated receptor disease/nonalcoholic steatohepatitis. Hepatol Res. 2015;45(4):363–77. https:// gamma; NAFLD: Nonalcoholic fatty liver disease; ELISA: Enzyme-linked doi.org/10.1111/hepr.12511. immunosorbent assay; IL-17: Interleukin (IL)-17; TNF: Tumor necrosis factor; 4. Bashiri A, Nesan D, Tavallaee G, Sue-Chue-Lam I, Chien K, Maguire GF, et al. IR: Insulin resistance; NASH: Nonalcoholic steatohepatitis; VLDL: very Low Cellular cholesterol accumulation modulates high fat high sucrose (HFHS) density lipoprotein; EMT: Epithelial-to-mesenchymal transition; IPGT diet-induced ER stress and hepatic inflammasome activation in the T: Intraperitoneal glucose tolerance test; IPITT: Intraperitoneal insulin development of non-alcoholic steatohepatitis. Biochim Biophys Acta. 2016; tolerance test; H&E: Hematoxylin and eosin; PBS: Phosphate-buffered saline;; 1861:594–605. RIPA: Radioimmunoprecipitation assay; GO: Gene Ontology; KEGG: Kyoto 5. Tilg H, Moschen AR. Evolution of inflammation in nonalcoholic fatty liver Encyclopedia of Genes and Genomes; BMI: Body mass index; disease: the multiple parallel hits hypothesis. Hepatology. 2010;52(5):1836– HOMA: Homeostasis model assessment; DE: Differentially expressed; 46. https://doi.org/10.1002/hep.24001. ACSL4: Acyl-CoA synthetase long chain family member 4; ANGP 6. Takaki A, Kawai D, Yamamoto K. Molecular mechanisms and new treatment TL4: Angiopoietin like 4; CPT1A: Carnitine palmitoyltransferase 1A; strategies for non-alcoholic steatohepatitis (NASH). Int J Mol Sci. 2014;15(5): PLIN4: Perilipin 4; HSCs: Hepatic stellate cells 7352–79. https://doi.org/10.3390/ijms15057352. Zhang et al. Lipids in Health and Disease (2021) 20:39 Page 13 of 13

7. Okubo H, Kushiyama A, Sakoda H, Nakatsu Y, Iizuka M, Taki N, et al. 25. Farrell GC, van Rooyen D, Gan L, Chitturi S. NASH is an inflammatory Involvement of resistin-like molecule beta in the development of disorder: pathogenic, prognostic and therapeutic implications. Gut Liver. methionine-choline deficient diet-induced non-alcoholic steatohepatitis in 2012;6(2):149–71. https://doi.org/10.5009/gnl.2012.6.2.149. mice. Sci Rep. 2016;6(1):20157. https://doi.org/10.1038/srep20157. 26. Del Campo JA, Gallego P, Grande L. Role of inflammatory response in liver 8. Petta S, Muratore C, Craxi A. Non-alcoholic fatty liver disease pathogenesis: diseases: therapeutic strategies. World J Hepatol. 2018;10(1):1–7. https://doi. the present and the future. Dig Liver Dis. 2009;41(9):615–25. https://doi. org/10.4254/wjh.v10.i1.1. org/10.1016/j.dld.2009.01.004. 27. Beringer A, Miossec P. IL-17 and IL-17-producing cells and liver diseases, 9. Gruben N, Shiri-Sverdlov R, Koonen DP, Hofker MH. Nonalcoholic fatty liver with focus on autoimmune liver diseases. Autoimmun Rev. 2018;17(12): disease: a main driver of insulin resistance or a dangerous liaison? Biochim 1176–85. https://doi.org/10.1016/j.autrev.2018.06.008. Biophys Acta. 2014;1842:2329–43. 28. Braunersreuther V, Viviani GL, Mach F, Montecucco F. Role of cytokines and 10. Hebbard L, George J. Animal models of nonalcoholic fatty liver disease. Nat chemokines in non-alcoholic fatty liver disease. World J Gastroenterol. 2012; Rev Gastroenterol Hepatol. 2011;8(1):35–44. https://doi.org/10.1038/nrga 18(8):727–35. https://doi.org/10.3748/wjg.v18.i8.727. stro.2010.191. 29. Kakino S, Ohki T, Nakayama H, Yuan X, Otabe S, Hashinaga T, et al. Pivotal 11. Kim JY, Lee C, Oh M, Im JA, Lee JW, Chu SH, et al. Relationship between role of TNF-alpha in the development and progression of nonalcoholic fatty non-alcoholic fatty liver disease, metabolic syndrome and insulin resistance liver disease in a murine model. Horm Metab Res. 2018;50:80–7. in Korean adults: a cross-sectional study. Clin Chim Acta. 2016;458:12–7. 30. Younossi ZM. Non-alcoholic fatty liver disease - a global public health https://doi.org/10.1016/j.cca.2016.03.018. perspective. J Hepatol. 2019;70(3):531–44. https://doi.org/10.1016/j.jhep.201 12. Colombo M, Raposo G, Thery C. Biogenesis, secretion, and intercellular 8.10.033. interactions of exosomes and other extracellular vesicles. Annu Rev Cell Dev 31. Chang CC, Wang SS, Hsieh HG, Lee WS, Chuang CL, Lin HC, et al. Biol. 2014;30(1):255–89. https://doi.org/10.1146/annurev-cellbio-101 Rosuvastatin improves hepatopulmonary syndrome through inhibition of 512-122326. inflammatory angiogenesis of lung. Clin Sci (Lond). 2015;129(6):449–60. 13. Hirsova P, Ibrahim SH, Krishnan A, Verma VK, Bronk SF, Werneburg NW, et al. https://doi.org/10.1042/CS20140622. Lipid-induced signaling causes release of inflammatory extracellular vesicles from hepatocytes. Gastroenterology. 2016;150(4):956–67. https://doi.org/10.1 Publisher’sNote 053/j.gastro.2015.12.037. Springer Nature remains neutral with regard to jurisdictional claims in 14. Ren R, Chen Z, Zhao X, Sun T, Zhang Y, Chen J, et al. A possible regulatory published maps and institutional affiliations. link between twist 1 and PPARgamma gene regulation in 3T3-L1 adipocytes. Lipids Health Dis. 2016;15(1):189. https://doi.org/10.1186/s12944- 016-0361-x. 15. Lu S, Wang H, Ren R, Shi X, Zhang Y, Ma W. Reduced expression of twist 1 is protective against insulin resistance of adipocytes and involves mitochondrial dysfunction. Sci Rep. 2018;8(1):12590. https://doi.org/10.1038/ s41598-018-30820-z. 16. Zhou L, Li Q, Chen A, Liu N, Chen N, Chen X, et al. via KLF15-activating Twist2 ameliorated hepatic steatosis by inhibiting inflammation and improving mitochondrial dysfunction NF-κB-FGF21 or SREBP1c-FGF21 pathway. FASEB J. 2019;33(12):14254–69. https://doi.org/10.1096/fj.201901347RR. 17. Franco HL, Casasnovas J, Rodriguez-Medina JR, Cadilla CL. Redundant or separate entities?--roles of Twist1 and Twist2 as molecular switches during gene transcription. Nucleic Acids Res. 2011;39(4):1177–86. https://doi.org/1 0.1093/nar/gkq890. 18. Merindol N, Riquet A, Szablewski V, Eliaou JF, Puisieux A, Bonnefoy N. The emerging role of twist proteins in hematopoietic cells and hematological malignancies. Blood Cancer J. 2014;4(4):e206. https://doi.org/10.1038/bcj.2 014.22. 19. Ansieau S, Bastid J, Doreau A, Morel AP, Bouchet BP, Thomas C, et al. Induction of EMT by twist proteins as a collateral effect of tumor-promoting inactivation of premature senescence. Cancer Cell. 2008;14(1):79–89. https:// doi.org/10.1016/j.ccr.2008.06.005. 20. Sosic D, Richardson JA, Yu K, Ornitz DM, Olson EN. Twist regulates cytokine gene expression through a negative feedback loop that represses NF- kappaB activity. Cell. 2003;112(2):169–80. https://doi.org/10.1016/S0092-8674 (03)00002-3. 21. Chen L, Chen R, Kemper S, Charrier A, Brigstock DR. Suppression of fibrogenic signaling in hepatic stellate cells by Twist1-dependent microRNA- 214 expression: role of exosomes in horizontal transfer of Twist1. Am J Physiol Gastrointest Liver Physiol. 2015;309(6):G491–9. https://doi.org/10.11 52/ajpgi.00140.2015. 22. Lallukka S, Sevastianova K, Perttila J, Hakkarainen A, Orho-Melander M, Lundbom N, et al. Adipose tissue is inflamed in NAFLD due to obesity but not in NAFLD due to genetic variation in PNPLA3. Diabetologia. 2013;56(4): 886–92. https://doi.org/10.1007/s00125-013-2829-9. 23. Moran-Salvador E, Lopez-Parra M, Garcia-Alonso V, Titos E, Martinez- Clemente M, Gonzalez-Periz A, et al. Role for PPARgamma in obesity- induced hepatic steatosis as determined by hepatocyte- and macrophage- specific conditional knockouts. FASEB J. 2011;25(8):2538–50. https://doi.org/1 0.1096/fj.10-173716. 24. Gavrilova O, Haluzik M, Matsusue K, Cutson JJ, Johnson L, Dietz KR, et al. Liver peroxisome proliferator-activated receptor gamma contributes to hepatic steatosis, triglyceride clearance, and regulation of body fat mass. J Biol Chem. 2003;278(36):34268–76. https://doi.org/10.1074/jbc.M300043200.